Deconstructing the Tower of London : resolution, alternatives and executive function

Abstract

Although widely employed as an executive task, the performance demands of the Tower of London task (TOL) are not well understood. The aim of this Thesis is to determine what TOL measures and seven studies are included that illustrate its performance demands. Much of the work is experimental and TOL is explored via a novel verification paradigm in which participants make speeded judgements about whether a demonstrated move is optimal or suboptimal. The deconstruction of the task to the level of the individual move indicated that the number of moves until a disk was placed in its goal position (resolution) and the number of legal alternative moves there were predicted performance. When the task was reconstructed to its original form where complete problems were issued, these performance demands continued to predict planning time, execution time and accuracy. These demands were characterised as involving a depth search (resolution) and a breadth search (alternatives) of the problem space. Several studies included individual difference measureso r the dual-taskp aradigmt o determinet he cognitive processesth at were involved in performance.It was argued that spatial processes are involved to manipulate disks in memory and it was argued that the resolution demands were closely related to planning and the alternative move demands to inhibition. These performance characteristics are involved in many real world tasks and may elicit the deficits observed in clinical samples. One study was presented in which the performance of older adults was compared to younger adults. It was shown that older adults had a particular deficit in their ability to consider alternative moves; a similar TOL deficit might be observed in other dysexecutive populations. This thesis provides recommendations and guidelines for using TOL in research and stresses the importance of the items used on performance.